Compatibility of Dynaco PAS with VTA ST70, Subwoofers, and other power amps -- INPUT IMPEDANCE discussion

I said write to him, not look up a web page.  Again, this thread is far afield.  But since you asked:

https://dynacotubeaudio.forumotion.com/t3109-vta-tube-buffer?highlight=Buffer

Audiobill, You have joined every thread I have started and crapped on it. Why do this? Can't you just ignore them? There is a lot of excellent technical discussion here by good folks who love Dynaco and love VTA. Who are you to judge my thread has gone "far afield"?

The Dynaco Tube Audio Forum
"Dedicated to the restoration and preservation of all original Dynaco tube audio equipment"

audiobill wrote:I said write to him, not look up a web page.  Again, this thread is far afield.  But since you asked:

https://dynacotubeaudio.forumotion.com/t3109-vta-tube-buffer?highlight=Buffer

This is actually useful. Looks like Roy sold a buffer in 2016. If it is still available I may buy one from him to use with my PH12 to protect the tape out loop. But it cannot be used with a PAS due to power supply limitations unfortunately. Hence, the reason we are looking at FETs.

The thread you posted includes the following link to a third party buffer that is apparently very similar

http://glass-ware.stores.yahoo.net/acpcbandusgu.html

.

rjpjnk wrote:

peterh wrote:You might eat your cake and keep it :-)
See this buffer that fits inside the PAS and may be powered by the PAS ( it's FET but so what ) :
https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-New/302849514977?hash=item46833cdde1:g:4a8AAOSwirFa95Pq

Peterh, Did you try this buffer yet? I found another one here that looks to be the same circuit.

https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-Buffer-New/292704780919

No. The snailmail is slow across the atlantic

Have you every seen just a bare board or a schematic for a FET buffer like these? Might be fun to build my own using choice parts.

Found this circuit from Google image search for example. Not sure how to evaluate its suitability. Can anyone identify this circuit?



EDIT: I found what appears to be the original source. Though the article doesn't say who designed the circuit or how well it works.
https://billthompson.us/bt/dyna-pas3-modifications/

UPDATE: I built the MOSFET buffer in the schematic posted above and it works great. I added two new output RCS jacks to the PAS so now I have the original output and the buffered output next to it. At present I am using the buffered output only to drive the subwoofer. I modified the circuit above slightly by changing the bias resistors from 470K to 1Meg so the input impedance of the buffer is 500K now. It measures great and sounds perfect. No more excessive loading cutting into the bass tone control response. It was only about 5 dollars in parts and a lot of fun to build. MOSFETS are pretty cool. I'd prefer tubes, but since the power supply is pretty limited thought I'd give it a try.

Spoke with the author, Bill Thompson, and he was very helpful, btw. Nice guy.

Here is a pic of the finished product on a breadboard.



And installed with two new jacks,

As Peter H suggests, an external buffer should bring matters dead to rights. To avoid impedance issues. , a dedicated subwoofer output jack is a very good thing to do. Relative to the work you have already completed, this is a small step. My buffer has variable volume, EQ, and a 3-pole switch to reverse phase and to shut off the subwoofer. I drive two subs to load the room, so this buffer, separate outs from the pre-amplifier has not only corrected the issues you’ve described, but yields great improvements.

If I did go external, I would build one with tubes, but for now the internal MOSFET buffer is definitely doing the trick. I have also now balanced the PAS to perfectly match the VTA ST-70 input impedance on the standard outputs, and am free to load the buffered outputs anyhow I like with no consequences. Woo Ho!

The PAS sounds great but it sure is a picky beast as far as loading. The stepped attenuator option on the ST-70 makes it a little less suitable for direct connection to a PAS preamp without modification. I am surprised this hasn't been discussed more, especially given that it is trivial to fix. I decided I like the attenuator, so I modified the PAS load resistors to keep the total load where Dynaco intended it to be. The result is excellent and the tone controls now work much better.

Now onto the next issue... The PAS phono stage is also unbuffered. This makes the tape loop next to useless with any modern equipment. I may experiment with buffering this as well. The reason is I would like to insert my Schiit Loki EQ in the tape loop and use that as an alternative to the PAS tone controls. However, the Loki has an input impedance of only 50K and the PAS phono stage is expecting about 225K. And it is *really* picky about its loading as well for the same reason as the line stage (tone controls in the NFB loop) Well, in this case the "tone control" is actually the RIAA filter, but same principle.

Using an inverse RIAA filter I swept the phono stage with and without the 50K load and found the bass rolloff is significant when the load is applied. It is down 6dB at 20Hz. Pretty much identical to the results published by audioregenesis.

So yeah, this PAS is a picky beast. Sounds great though when it is happy.

peterh wrote:

rjpjnk wrote:

peterh wrote:You might eat your cake and keep it :-)
See this buffer that fits inside the PAS and may be powered by the PAS ( it's FET but so what ) :
https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-New/302849514977?hash=item46833cdde1:g:4a8AAOSwirFa95Pq

Peterh, Did you try this buffer yet? I found another one here that looks to be the same circuit.

https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-Buffer-New/292704780919

No. The snailmail is slow across the atlantic

I have now received the buffer. And yes, it fits fine and should work fine.
I do not however have any schematics , the seller refuses thus i will do
some reverse engineering and publish the findings.

I have no plans using it in my own set as i already have made buffers for
both audio out and tape out. Purchase was for evaluation for possible resale.
( no i will not copy this, i intend to purchase from the seller, but i refuse
to sell anything without proper documentation)

Note: I no longer need a buffer for the phono stage because I was able to modify the Schiit Loki EQ input impedance.

In the mood to explore, I opened up the Loki to have a look at its circuits and saw that that it used a 475R resistor followed by a 47.5K shunt resistor to ground at the input to the first SS device (transistor?). I swapped the 47.5K resistor with a 220K and now the input impedance is 220K and the performance seems unchanged. Other than the fact that these components are extremely tiny SMDs it was a simple mod.

The Loki now seems to work perfectly in the PAS 3x tape loop. It now presents the impedance the phono stage is expecting and hence no longer cuts the bass response 6dB when I connect it.

I wonder what the consequences are...

Interesting. Can you post pics showing which resistor to change? I also have a Loki EQ and would like to buffer the phono out on my PH12 which is built into my SP13.

wgallupe wrote:Interesting. Can you post pics showing which resistor to change? I also have a Loki EQ and would like to buffer the phono out on my  PH12 which is built into my SP13.

Sure, I will get a picture tonight and post it. Good point about the PH12 phono stage. It is also unbuffered I know. What impedance is the PH12 able to accommodate on its outputs without degradation of performance?

rjpjnk wrote:

wgallupe wrote:Interesting. Can you post pics showing which resistor to change? I also have a Loki EQ and would like to buffer the phono out on my  PH12 which is built into my SP13.

Sure, I will get a picture tonight and post it. Good point about the PH12 phono stage. It is also unbuffered I know. What impedance is the PH12 able to accommodate on its outputs without degradation of performance?

I don't know but there's a thread here somewhere that suggests 100K minimum is good. Thanks in advance for the pics!

Here is the picture of the Loki with one channel modified. The arrow points to the original 47.5K resistor that gets changed to 200-220K. In this pic I have changed only channel 1. You can see the huge 1/2 watt resistor soldered in place where the tiny SMD used to be. Channel 2 is untouched in this pic.

Excellent! Thanks for posting that. Couple questions: Is a 1/2 watt resistor really needed? Any problems un-soldering the SMD? Or inserting the relatively large resistor leads into the holes on the PCB?

No, the 1/2 watt resistor is not needed. It is just there as an experiment. It is just all I had laying around at the time. I have since ordered some 2mm SMD resistors for pretty.

That said, there is no reason you can't use a 1/2 or 1/4 watt axial resistor like in the picture if you want. There is plenty of room and they are easier to solder than the SMDs. Regardless of the choice, the leads are soldered on top of the circuit board. There are no holes.

You need an extremely tiny tip soldering iron and a small flat jewler's screwdriver or tweezers to lift the SMD. And a magnifying glass.

peterh wrote:

peterh wrote:

rjpjnk wrote:

peterh wrote:You might eat your cake and keep it :-)
See this buffer that fits inside the PAS and may be powered by the PAS ( it's FET but so what ) :
https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-New/302849514977?hash=item46833cdde1:g:4a8AAOSwirFa95Pq

Peterh, Did you try this buffer yet? I found another one here that looks to be the same circuit.

https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-Buffer-New/292704780919

No. The snailmail is slow across the atlantic

I have now received the buffer. And yes, it fits fine and should work fine.
I do not however have any schematics , the seller refuses thus i will do
some reverse engineering and publish the findings.

I have no plans using it in my own set as i already have made buffers for
both audio out and tape out. Purchase was for evaluation for possible resale.
( no i will not copy this, i intend to purchase from the seller, but i refuse
to sell anything without proper documentation)

Schematic done on paper. Will transfer to printable form later, but no surprises here.
And i will probably test it in a extra PAS3X i have

Looking forward to your results.
How will you evaluate it?

rjpjnk wrote:Looking forward to your results.
How will you evaluate it?

Not decided yet. Might be a simple listering test, maybe adding some load.
The nice thing with this design is that it is driven by symmetrical power ( +12v and -12V)
thus there is no cap charge hump at power on or off. Given that the pilot light is
replaced by a LED the total load on the transformer is lower then stock version.

rjpjnk wrote:... I have since ordered some 2mm SMD resistors for pretty.

Soldered in the tiny 220K SMDs last night and it looks like new again. I taped the original resistors to the inside case in case I ever want to revert back to them. Did a lot of testing of the frequency response and the modified Loki seems to work well with the PAS now.

One thing to keep in mind though. If anything is plugged into the Tape Out jack and the Tape/Input switch is in the "Input" position, then the impedance of the device attached to the Tape Out appears in parallel to the normal impedance of the Line Stage (That is, vol control, balance control, loudness tap, and 47K series resistor). Maybe this is obvious to many, but some folks may not realize this and it is important because leaving this device plugged in will impact the bass response even though the device is not being used.

For example, even the modified Loki cannot be left plugged into the tape loop if the Tape/Input switch is in the "Input" position because then the impedance seen by the Phono Stage would be only 110K = (220 Loki) || (220 PAS). It is not a huge problem, and certainly much better than when the Loki was at 50K, but even this small change does impact the bass response.

FET buffer installed : http://n.manet.nu/pas3/PAS-buffer.jpg

Or the whole amp:
http://n.manet.nu/pas3/index.html    Scroll down to last picture

First impression : sounds ok, no harsh flutes, no noice.
As for mounting : the metal angles used some strange threds. A M3 threadmaker ( what that is called
in english ) fixed this.
The mounting is tight, in aftersight one should solder all wires on the buffer first, then install and
last solder wires into the PAS at right positions.

I found a similar item from hongkong :
https://www.ebay.com/itm/Dynaco-PAS-Preamp-Impedance-Matching-Module-Buffer-New-/292704780919?oid=302723561677

Seems to be a straight ripoff at a lower price.

As a comment,
Mounting 2 of these boards would be hard,
a buffer with 4 channels could be nice, this would make the tape-out usable.

Well done. Looking forward to the schematic.

I looked at that Chinese version as well, and spoke with the seller a couple times to inquire about buying unpopulated boards. He would sell 2 for 9.00 + shipping. Problem is the schematic calls for 2SK170 JFETS. These are classic, famous Toshiba parts from back in the 70's I believe, and they are no longer in production. NOS are 20-60 dollars each on ebay. Of course, the Chinese have imitations available for 50 cents. These are the ones included one the board. After more research I found the market is flooded with these fake JFETS with many claiming to be original, but all are actually Chinese rip-offs just marked up sufficiently to make buyers think they are genuine. Sad.

That's why I built the other circuit.

Anyway, your whole amp look fantastic. Excellent work!

rjpjnk wrote:Well done. Looking forward to the schematic.

I looked at that Chinese version as well, and spoke with the seller a couple times to inquire about buying  unpopulated boards. He would sell 2 for 9.00 + shipping. Problem is the schematic calls for 2SK170 JFETS. These are classic, famous Toshiba parts from back in the 70's I believe, and they are no longer in production. NOS are 20-60 dollars each on ebay. Of course, the Chinese have imitations available for 50 cents. These are the ones included one the board. After more research I found the market is flooded with these fake JFETS with many claiming to be original, but all are actually Chinese rip-offs just marked up sufficiently to make buyers think they are genuine. Sad.

That's why I built the other circuit.

Anyway, your whole amp look fantastic. Excellent work!

Yes, it's sk170 here too. In a very classical source follower, using another sk170 as
current source ( source and gate connected together, drain connects to source of
source-follower)
I made a schematic on papers, i'll try to upload it.
( upload failed, try : )
http://n.manet.nu/pas3/FET-buffer.jpg

My dedicated buffer has been overhauled :
http://n.manet.nu/fm1/second.html

by installing proper RCA connectors and a general fixup

To summarize, this former fm1 has 2 pairs of cathode buffers, both for the "amp channel" but
more importent for me, a buffered tape-out channel to allow computer soundcard to be
connected. In addition it contains a remote capable volume control and a power sequencer
to turn on/off pre and power amp, protectimng from power outages.